Bilateral intercomparison of the calibration of type S thermocouples in the fixed points of Zn, Al, Ag and Au

Project Description

Two thermocouples type S, Manufactured at GUM were first calibrated at GUM, thereafter at NMi-VSL and finally at GUM in the fixed points Zn, Al, Ag and Au. Measurements, uncertainty budgets and methods are compared.

Final Report 2003-02-03

Two thermocouples type S, Manufactured at GUM (Id JM50, JM51) were first calibrated at GUM, thereafter at NMi-VSL and finally at GUM in the fixed points Zn, Al, Ag and Au. Measurements, uncertainty budgets and methods are compared. Relevant information concerning the equipment is listed in table 1. The results are summarised in table 2 an example of an uncertainty budget is shown in table 3. In Figure 1 the result of the measurements on the two thermocouples are shown as the differences: EGUM- ENMi with the expanded combined uncertainty. The uncertainty is dominated by the contribution from the in-homogeneity of the thermocouple.
At GUM two thermocouples were manufactured using 0,5 mm diameter Pt and Pt-Rh wires (Johnson Matthey). The thermocouple wire (1300 mm in length) was electrically annealed in air at 1150 °C for 1 h. The thermocouples were mounted in twin-bore alumina tubes and protection alumina tubes closed at one end both with 99,7 % purity. The mounted thermocouples were annealed in a furnace at 1100 °C for 4 h. During measurement the reference junction of each thermocouple was realized at 0 °C by using a vacuum flask with a mixture of shaved ice and distilled water. The thermocouples were calibrated at the fixed points of Zn, Al, Ag, Au. The fixed point plateaus were realised by first melting the metal ingot completely by keeping it at a temperature about 10 K above the freezing point for about 20 minutes. The temperature of the furnace was the set 1 K below the freezing temperature. The calibration of thermocouples was performed about 20% of the total plateau length. For each freezing point plateau at least two repeat measurements were made.
At NMi-VSL the thermocouples were initially annealed at 1050 °C for on hour. They were then calibrated at the fixed points Zn, Al, Ag, Au. At each freezing point plateau, three repeat measurements were made. A check on the thermocouples homogeneity for use in the uncertainty calculation was made at each fixed point by extracting the thermocouple five centimetres, noting the change in emf.
After the comparison a second set of measurements were made at GUM to check the immersion characteristics (homogeneity) of the two thermocouples in the GUM fixed points. The last measurements with JM51 were not included because of drift during the calibration.